TAUR-dev/Q7BI_reflection__countdown_3arg__reflection_chunk_18

该数据集包含了问题(question)、答案(answer)、任务配置(task_config)、任务来源(task_source)、提示(prompt)、元数据(metadata)、模型响应是否正确(is_model_response_correct)、模型响应的推理(is_model_response_correct__reasoning)、示例注释(prompt__example_annotation__metadata)、模型响应(model_responses__example_annotation)、提取的答案(model_responses__example_annotation__eval_extracted_answers)、提示类型(prompt_type)、反思提示(reflection_prompt)、变异提示的反思(model_responses__mutated_prompts_reflection)等信息。数据集适用于训练模型进行问答等任务。

The dataset includes fields such as question, answer, task configuration, task source, prompt, metadata, whether the models response is correct, the reasoning of the models response, example annotation metadata, model response, extracted answers from the model response, prompt type, reflection prompt, reflection on mutated prompts, and other information. The dataset is suitable for training models for tasks such as question answering.